[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [modeller_usage] Converting (old) top script to (new) python ones



Eric Hajjar wrote:
The new "polished - python" interface looks really promising , but i'm spending now some time to see how to make things work !
I went throught all tutorials , and that is OK ,
but in my present case i want to add DISULFIDE RESTRAINTS and i want to change the default "coordinate-output-names"
so if i understand right i have to NOT use the "automodel" class anymore...

Here starts the trouble !

Also: Before i was doing :
SUBROUTINE ROUTINE = 'special_patches'
PATCH RESIDUE_TYPE = 'DISU', RESIDUE_IDS = '27' '50'
RETURN
END_SUBROUTINE

To do this with the new Python interface does not require you to do anything difficult - just like before, you simply redefine the 'special_patches' routine. See http://salilab.org/modeller/FAQ.html#3

I'm not sure what you mean by 'change the default coordinate-output-names'. If you mean that you don't like the .B99*.pdb naming that Modeller uses, there's no easy way to do that in 8v1 (although there will be in the next release). The easiest thing to do would probably be to redefine the single_model routine. See modlib/modeller/automodel/automodel.py for the original version of the routine. Of course, there's no easy way to do this with TOP either.

now , according to the manual :
"mdl.patch_ss_templates(aln)
# Create the stereochemical restraints
mdl.restraints.make(aln, restraint_type='stereo', spline_on_site=False) "


If i follow those examples in the manual and try to read my alignement to write the correponding model-with patches constraints. i get a completely unfolded, linear protein as a result !

The example you give does not build a model, but simply builds an approximate model from the template and internal coordinates. There are many more steps required to build a comparative model (for example, you need to build distance restraints to restrain the model to be like the template).

	Ben Webb, Modeller Caretaker
--
             http://www.salilab.org/modeller/
Modeller mail list: http://salilab.org/mailman/listinfo/modeller_usage